Download Antifungal Activity of Amiodarone Is Mediated by Disruption of

THE JOURNAL OF BIOLOGICAL CHEMISTRY
© 2003 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 278, No. 31, Issue of August 1, pp. 28831–28839, 2003
Printed in U.S.A.
Antifungal Activity of Amiodarone Is Mediated by Disruption of
Calcium Homeostasis*
Received for publication, March 31, 2003, and in revised form, May 7, 2003
Published, JBC Papers in Press, May 16, 2003, DOI 10.1074/jbc.M303300200
Soma Sen Gupta‡§¶, Van-Khue Ton‡¶, Veronica Beaudry储, Samuel Rulli‡, Kyle Cunningham储,
and Rajini Rao‡**
From the Departments of ‡Physiology and 储Biology, The Johns Hopkins University, Baltimore, Maryland 21205
The antiarrhythmic drug amiodarone was recently
demonstrated to have novel broad range fungicidal activity. We provide evidence that amiodarone toxicity is
mediated by disruption of Ca2ⴙ homeostasis in Saccharomyces cerevisiae. In mutants lacking calcineurin and
various Ca2ⴙ transporters, including pumps (Pmr1 and
Pmc1), channels (Cch1/Mid1 and Yvc1), and exchangers
(Vcx1), amiodarone sensitivity correlates with cytoplasmic calcium overload. Measurements of cytosolic Ca2ⴙ
by aequorin luminescence demonstrate a biphasic response to amiodarone. An immediate and extensive calcium influx was observed that was dose-dependent and
correlated with drug sensitivity. The second phase consisted of a sustained release of calcium from the vacuole
via the calcium channel Yvc1 and was independent of
extracellular Ca2ⴙ entry. To uncover additional cellular
pathways involved in amiodarone sensitivity, we conducted a genome-wide screen of nearly 5000 single-gene
yeast deletion mutants. 36 yeast strains with amiodarone hypersensitivity were identified, including mutants in transporters (pmr1, pdr5, and vacuolar HⴙATPase), ergosterol biosynthesis (erg3, erg6, and erg24),
intracellular trafficking (vps45 and rcy1), and signaling
(ypk1 and ptc1). Of three mutants examined (vps45,
vma3, and rcy1), all were found to have defective calcium homeostasis, supporting a correlation with amiodarone hypersensitivity. We show that low doses of
amiodarone and an azole (miconazole, fluconazole) are
strongly synergistic and exhibit potent fungicidal effects in combination. Our findings point to the potentially effective application of amiodarone as a novel
antimycotic, particularly in combination with conventional antifungals.
The emergence of drug-resistant fungi poses an increasing
threat to the treatment of opportunistic fungal infections in
patients with compromised immune systems, as occurs commonly in cancer and AIDS. This can only be countered by the
discovery of new antifungal agents, particularly those that
* This work was funded by a Burroughs Wellcome Student Elective
Prize (to S. S. G.) and by National Institutes of Health Grant GM62142
and a grant-in-aid from the American Heart Association Mid-Atlantic
Affiliate (to R. R.). The costs of publication of this article were defrayed
in part by the payment of page charges. This article must therefore be
hereby marked “advertisement” in accordance with 18 U.S.C. Section
1734 solely to indicate this fact.
§ Present address: Lucy Cavendish College, University of Cambridge,
Cambridge CB3 OBU, United Kingdom.
¶ These authors contributed equally to this work.
** To whom correspondence should be addressed: Dept. of Physiology,
The Johns Hopkins University School of Medicine, 725 N. Wolfe St.,
Baltimore, MD 21205. Tel.: 410-955-4732; Fax: 410-955-0461; E-mail:
[email protected].
This paper is available on line at http://www.jbc.org
target different molecular pathways, and a better understanding of their mode of action. Recently, the antiarrhythmic drug
amiodarone (AMD)1 was shown to have potent fungicidal activity against not only Saccharomyces but also pathogenic
yeasts such as Candida, Cryptococcus, Fusarium, and Aspergillus (1). Viability of Cryptococcus neoformans after exposure
to 10 ␮M AMD fell dramatically, with a 10-fold loss of viable
cells after 1 h of treatment and only 0.001% viable cells remaining after 1 day. The cytotoxic effect of AMD appeared to be
mediated by calcium, as evidenced by the ability of high (millimolar) concentrations of extracellular Ca2⫹ to ameliorate
drug toxicity (1) and by an increase in cytosolic Ca2⫹ upon drug
application (2).
There is emerging evidence that cytosolic Ca2⫹ entry in yeast
is critical for survival under a variety of cell stresses, including
hyper- and hypo-osmotic shock, protein unfolding agents, and
antifungal drugs (3– 6). The release of calcium from intracellular stores must be compensated by stimulation of extracellular
calcium influx, a phenomenon commonly known as capacitative
calcium entry (7). Thus, one possibility is that the cytosolic
Ca2⫹ increase in response to AMD promotes cell survival and is
due to capacitative calcium entry, as has been suggested by
Courchesne and Ozturk (2). Paradoxically, excessive or unregulated levels of calcium in the cytoplasm also lead to cell death
and are implicated in the cytotoxicity of several drugs (8) as
well as fungal toxins (9, 10). To better understand the role of
calcium in AMD toxicity, it is important to distinguish between
a capacitative calcium entry mechanism triggered by store
depletion or a loss in calcium homeostasis, possibly due to an
effect of the drug on ion channels and transporters. The experiments described in this work address these two possibilities
and begin to elucidate the cellular basis of AMD toxicity.
We present a detailed investigation of how AMD disrupts
Ca2⫹ homeostasis in Saccharomyces cerevisiae. The roles of
different Ca2⫹ transporters, including pumps (Pmr1 and
Pmc1), channels (Cch1, Mid1, and Yvc1), and exchangers
(Vcx1), were evaluated in cells lacking these proteins and exposed to relatively low, therapeutically relevant levels of AMD.
Pmr1 is an ATP-driven pump that sequesters Ca2⫹ and Mn2⫹
into the Golgi/secretory pathway and maintains cellular ion
homeostasis under normal growing conditions (11, 12). Homologues of Pmr1 are widely distributed in organisms including
Caenorhabditis elegans, Drosophila, and humans and constitute the newly recognized SPCA subtype of Ca2⫹-ATPases (13).
Pmc1 is the yeast Ca2⫹ pump related to mammalian plasma
membrane Ca2⫹-ATPases that is induced under calcium stress
(14) or in the absence of Pmr1 (12) and serves to detoxify excess
1
The abbreviations used are: AMD, amiodarone; BAPTA, 1,2-bis(2aminophenoxy)-ethane-N,N,N⬘,N⬘-tetracetic acid; FLUC, fluconazole;
MIC, miconazole; RLU, relative luminescence unit(s); WT, wild type;
SC, synthetic complete.
28831
28832
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
Ca2⫹ by sequestration into the vacuole. Vacuolar calcium sequestration is also accomplished by the H⫹/Ca2⫹ exchanger,
Vcx1 (15–17), which requires the proton electrochemical gradient generated by the vacuolar H⫹-ATPase. Calcium release
from the vacuole is mediated by Yvc1, a transient receptor
potential-like Ca2⫹ channel (18). Cch1 is the yeast homologue
of mammalian voltage-gated calcium channels and, together
with the stretch-activated Ca2⫹ channel Mid1, constitutes a
Ca2⫹ entry channel at the plasma membrane (19). There is
experimental evidence for at least one other calcium influx
channel of unknown molecular identity (5, 20). Together, these
proteins regulate cellular calcium levels and are critical for
proper functioning of the calcium signaling cascade. Calcium
activation of calmodulin and the consequent activation of the
calcium- and calmodulin-activated protein phosphatase, calcineurin, are believed to lead to a number of transcriptional
and post-translational signals that mediate a variety of different cellular responses to extracellular cues, including cell cycle
progression and the protein kinase C-mediated cell wall integrity pathway (reviewed in Refs. 21 and 22). The conservation of
several components of the calcium homeostatic machinery with
higher eukaryotes makes yeast an excellent model to study the
role of calcium signaling in AMD toxicity.
The results of our analysis of deletion mutants point to a
requirement for Pmr1-mediated Ca2⫹ homeostasis in the
growth response to AMD. Similar to recent observations on
other drug resistance mechanisms (3, 6), we also demonstrate
a critical role for calcineurin in AMD tolerance. Analyses of
cytoplasmic Ca2⫹ levels by aequorin luminescence assays indicate a correlation between AMD-induced cytosolic Ca2⫹ increase and growth toxicity. Specifically, AMD seems to cause
both Ca2⫹ influx at the cell membrane and release from internal stores, including the vacuole. To identify additional factors
that mediate drug sensitivity, we screened nearly 5000 singlegene deletion mutants for hypersensitivity to AMD. We identify 36 mutants, implicated in known and novel pathways that
may be important for drug resistance and detoxification. The
results of this study indicate that AMD kills yeast by a mechanism different from that used by conventional antifungals
such as the azoles and the polyenes. Finally, we show that low
concentrations of AMD and an azole (miconazole and fluconazole) are strikingly synergistic in combination, suggesting
that supplementation of conventional antifungal treatment
with AMD may be a practical means of converting the fungistatic effect of the azoles into more effective fungicides. In
summary, our findings provide important insights into the
cellular effects of AMD and its potential efficacy in the treatment of fungal infections.
EXPERIMENTAL PROCEDURES
Yeast Strains and Media—Isogenic sets of yeast deletions were all
derived from W303 and have been described before (15). The MAT␣
S. cerevisiae deletion library was purchased from Research Genetics
(Invitrogen). The Q783A and D53A mutations of PMR1 have been
described (23, 24). Candida albicans SC5314 (25) was obtained from the
laboratory of Brendan Cormack at The Johns Hopkins University
School of Medicine, and C. neoformans (JEC21) (26) was a gift from the
laboratory of Joseph Heitman at Duke University Medical Center. All
yeast strains were grown in synthetic complete (SC) medium (Bio 101,
Inc., Vista, CA) or standard YPD medium (2% Difco yeast extract, 1%
bacto-peptone, 2% dextrose) at 30 or 37 °C as specified in the legends
to Figs. 1, 5, and 6.
Growth Assays for Drug Sensitivity—AMD was purchased from
Sigma, dissolved in dimethyl sulfoxide as a 20 mM stock, and stored at
⫺20 °C. 10 ␮l of saturated seed cultures were inoculated in 3 ml of
growth medium containing different concentrations of AMD (0 –15 ␮M).
Cultures were grown in a 30 °C shaker for 24 h, and growth was
determined by optical density at 600 nm. To test the effect of divalent
cations, AMD sensitivity was monitored in the presence of 10 –25 mM
CaCl2 or 10 mM MgCl2. Alternatively, the cation chelator 1,2-bis(2-
aminophenoxy)-ethane-N,N,N⬘,N⬘-tetracetic acid (BAPTA; Sigma) was
added to final concentrations of 0.5 and 2 mM. To examine the role of
calcineurin inhibition, FK506 was diluted from a stock of 0.2 mg/ml to
a final concentration of 1 ␮g/ml and added to cell cultures grown in
stationary 96-well plates at 30 °C.
Miconazole and fluconazole were diluted from a stock of 5 mM (in
Me2SO) or 2 mg/ml (in water), respectively. In the fluconazole screen
with C. albicans, the seed culture was diluted by 1:100 and added to 1
ml of YPD containing different drug concentrations in a 24-well plate.
After a 24 – 48-h incubation at 37 °C, 5 ␮l of cells were diluted into 1.5
ml of YPD containing no drugs and grown overnight. Growth was
measured by absorbance at 600 nM. With C. neoformans, 50 ␮l of
saturated seed culture was added to 5 ml of YPD supplemented with
various drug concentrations. Cells were shaken vigorously for 20 h at
30 °C. 5 ␮l of cells exposed to AMD alone or 20 ␮l of cells exposed to both
drugs were then diluted in 10 ml YPD and grown to saturation for
24 – 48 h. Growth was measured by absorbance at 600 nM.
Genome-wide AMD Sensitivity Screen—200 ␮l of SC medium supplemented with 7 ␮M AMD was added to a 96-well plate and inoculated
with 5 ␮l of freshly grown stationary phase yeast culture derived from
single-gene deletion strains from the ResGen MAT␣ deletion library
(Invitrogen). Control cultures contained an equal volume of dimethyl
sulfoxide. The plates were incubated at room temperature for a period
of 12–24 h. Cultures were resuspended with a multichannel pipettor,
and growth was monitored by measuring the absorbance at 600 nm in
a SPECTRAmax 340 microplate reader (Molecular Devices). The relative growth of each strain was expressed as a percentage of A600 of the
control culture (i.e. no AMD). Strains with enhanced sensitivity to AMD
were selected based on growth inhibition of 80% or greater relative to
control. All candidate strains were retested for enhanced sensitivity in
4 and 8 ␮M AMD.
Aequorin Luminescence Assay—Yeast strains were transformed with
plasmid pEVP11-Aeq-89 (4) carrying the aequorin gene and then grown
to midlog phase in SC medium lacking leucine. Cells were harvested,
incubated with 0.25 mg/ml coelenterazine (Molecular Probes, Inc.,
Eugene, OR) in the dark for 20 min, washed and resuspended in
medium, and then incubated at 30 °C for 90 min to allow recovery. AMD
(at specified concentrations) was added to 0.3 ml of cells in a cuvette
through an injector, and luminescence was immediately recorded every
second for 10 min in a luminometer (Lumat LB 9507). The maximal
luminescence (Lmax) was determined as described (26). Specifically,
cells were lysed with 1% digitonin in the presence of 1 M CaCl2, and a
peak of RLU was recorded. Calcium concentration was calculated with
Equation 1,
[Ca2⫹]⫽ ((L/Lmax)1/3 ⫹ [118(L/Lmax)1/3] ⫺ 1)/
(7 ⫻ 106 ⫺ [7 ⫻ 106(L/Lmax)1/3])
(Eq. 1)
where L represents the RLU at a given time point. The base-line
calcium concentration of pmc1 in SC and Me2SO was ⬃0.38 ␮M, and
that of pmr1 was about 0.52 ␮M.
Methylene Blue Viability Assay—S. cerevisiae cells grown in media
with and without the specified concentrations of AMD and miconazole
were stained with methylene blue (Sigma) taken from a stock of 0.1
mg/ml. 250 –300 cells were counted under a light microscope. Dead cells
stain blue.
FUN-1 Confocal Microscopy—The Live/Dead yeast viability kit was
purchased from Molecular Probes. 50 ␮l of cells grown overnight in the
presence or absence of drugs (as specified in Fig. 5) were incubated in
SC medium at 30 °C for 1 h with 4 ␮M FUN-1 dye, which was diluted in
SC from a stock of 200 ␮M in Me2SO. Immediately after incubation, the
cells were examined under a confocal laser-scanning microscope
(PerkinElmer UltraView LCI System) equipped with an inverted ⫻ 100
oil immersion objective lens. The fluorescent dye was excited at 488 nm
by the krypton/argon laser. Conversion of FUN-1 into cylindrical intravacuolar structures was monitored by recording fluorescent micrographs at emission wavelengths of 600 nm (metabolically active and
inactive cells) or 520 nm (metabolically inactive cells only). Pseudocolorization was done with Adobe Photoshop software (Adobe Systems Inc.).
RESULTS
Hypersensitivity of the pmr1 Mutant to Amiodarone—To investigate the role of various calcium transport pathways in
mediating drug sensitivity, we compared the growth of individual, isogenic gene deletion mutants in media supplemented
with micromolar concentrations of AMD. As shown in Fig. 1A,
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
28833
FIG. 1. AMD sensitivity is modulated by calcium. A, Ca2⫹ pump mutants have differential sensitivities to AMD. Wild-type K601 (WT) and
isogenic pmc1 and pmr1 mutants were inoculated into 3 ml of SC media supplemented with AMD as shown and grown for 24 h at 30 °C. Growth
(optical density at 600 nm) is depicted as percentage of the control culture without drug for each strain. Data points are averages of duplicates,
with S.D. values within 5%. One experiment of three is shown. B, divalent cations protect against AMD toxicity. Wild-type K601 (WT) cells were
grown in AMD-containing SC medium supplemented with one of the following: CaCl2 (10 mM), MgCl2 (10 mM), or BAPTA (0.5 mM or 2 mM) at 30 °C
for 24 h. Growth was monitored as described above. C, ion selectivity mutants of Pmr1 have differential sensitivities to AMD. Strain K616
(pmr1pmc1cnb1) was transformed with a 2␮ plasmid carrying wild type Pmr1 or mutants D53A or Q783A and then grown in AMD-supplemented
SC medium lacking uracil as described above. Wild type Pmr1 transports both Ca2⫹ and Mn2⫹ equally well, whereas mutant D53A is defective in
Ca2⫹ transport, and mutant Q783A is defective in Mn2⫹ transport. D, loss of calcineurin increases AMD sensitivity. Pairs of isogenic wild type
(K601 or BY4742) and mutant strains (pmr1 and cnb1) were grown in SC media containing 4 or 8 ␮M AMD, with FK506 (1 ␮g/ml) added as
indicated. Growth was monitored as described above.
the pmr1 mutant was strikingly more sensitive to AMD; in
contrast, the pmc1 null strain showed a small but reproducible
tolerance to the drug, relative to wild type. A survey of deletion
mutants of the vacuolar H⫹/Ca2⫹ exchanger (vcx1) and all
known Ca2⫹ channels in yeast (yvc1, cch1, and mid1) revealed
a modest exacerbation of AMD sensitivity in both cch1 and
mid1 mutants but not in the others (not shown). Taken together, these results suggest that some aspect of Ca2⫹ signaling, maintained in large part by the ATP-dependent calcium
efflux pump, Pmr1, is important in mediating the growth toxicity of AMD.
Extracellular Calcium Modulates Amiodarone Toxicity—Extracellular calcium has been reported to confer a dose-dependent abrogation of susceptibility to KP4 fungal toxin in the yeast
Ustilago maydis (28). Here, 10 mM CaCl2 or MgCl2 appeared to
protect cell growth against AMD toxicity, whereas BAPTA, a
membrane-impermeant cation chelator, enhanced AMD toxicity, consistent with the removal of Ca2⫹ (Fig. 1B). Interestingly, the protective effect of increasing Ca2⫹ concentrations
(10 –25 mM) on AMD sensitivity was most significant in strains
lacking Pmr1 (not shown), which may reflect, at least in part,
the known dependence of this strain on extracellular Ca2⫹ for
optimum growth (29). Mn2⫹ can be a surrogate for Ca2⫹ in a
number of physiological processes but was ineffective in conferring protection (data not shown). Monovalent ions (K⫹ and
Na⫹) were previously reported to have little or no effect on
AMD toxicity to yeast (1).
Ion Selectivity Mutants of Pmr1 Confirm the Specific Role of
Calcium in Amiodarone Toxicity—Since Pmr1 mediates the
high affinity transport of both Ca2⫹ and Mn2⫹ into the Golgi
(11, 24), it was of interest to determine whether the enhanced
sensitivity of pmr1 mutant to AMD was a consequence of a loss
of Ca2⫹ or Mn2⫹ transport. We have previously described two
ion selectivity mutants of Pmr1, Q783A and D53A, which show
nearly exclusive transport of either Ca2⫹ or Mn2⫹, respectively
(23, 24). These mutants were expressed in a yeast strain devoid
of endogenous Ca2⫹ pumps, with the additional deletion of
calcineurin to maintain viability (pmr1pmc1cnb1) (11, 15). Selective loss of Ca2⫹ transport in Pmr1 mutant D53A resulted in
loss of tolerance to AMD, similar to the host strain lacking
28834
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
FIG. 2. AMD-induced cytoplasmic Ca2ⴙ elevation is biphasic and elevated in the pmr1 mutant. Ca2⫹-dependent luminescence of the
aequorin-coelenterazine photoprotein complex expressed in the cytosol of WT (A), pmc1 (B), and pmr1 (C) mutants was measured upon the addition
of AMD (8 –20 ␮M) to cells resuspended in SC medium at 1 OD/ml, as described under “Experimental Procedures.” Symbols corresponding to the
AMD concentrations are shown in the inset to A. Luminescence (RLU) is measured in arbitrary units/s, over 10 min. Calibration of RLU to free
calcium concentration is described under “Experimental Procedures.” D, a comparison of luminescence in response to 10 ␮M AMD. The pmr1
mutant shows greater elevations of aequorin luminescence, relative to WT and pmc1.
Pmr1 and Pmc1 (Fig. 1C). In contrast, selective loss of Mn2⫹
transport of mutant Q783A did not alter AMD sensitivity.
These results are consistent with the inability of extracellular
Mn2⫹ to protect against AMD toxicity, and they establish the
specific role of Ca2⫹ ions in mediating the cellular effects of
the drug.
Loss of Calcineurin Function Enhances Toxicity of Amiodarone—Calcineurin has been shown to be nonessential for normal growth but critical for survival during membrane stress in
C. albicans (3). The enhanced AMD sensitivity of the triple
mutant, pmr1pmc1cnb1 (Fig. 1C) relative to the individual
gene deletions (Fig. 1A) suggested that loss of calcineurin may
be synergistic with loss of the Golgi Ca2⫹ pump Pmr1. To
confirm this, we investigated the effect of FK506 (1 ␮g/ml), a
potent inhibitor of calcineurin, on the AMD sensitivity of wild
type and pmr1 strains (Fig. 1D). As predicted, the addition of
FK506 exacerbated the AMD sensitivity of the pmr1 mutant.
Additionally, the cnb1 mutant, lacking functional calcineurin,
displayed increased sensitivity to AMD at higher concentrations of the drug (Fig. 1D). Taken together, these data confirm
the importance of calcineurin in the Ca2⫹ signaling pathway
involved in drug sensitivity.
Amiodarone Triggers Calcium Influx in Yeast Cells—Next,
we investigated whether AMD induced calcium entry into yeast
and whether calcium entry correlated with drug sensitivity.
Calcium-dependent luminescence of the aequorin-coelenterazine photoprotein complex was monitored in the first few minutes immediately following drug addition. In all strains examined, application of AMD resulted in a biphasic elevation of
cytoplasmic calcium that was dependent on AMD concentration (Fig. 2, A–C). The first peak of calcium occurred extremely
fast (within 1–1.5 min) and was followed by a sustained rise
that lasted for the duration of the assay. The magnitude of both
phases, as well as the kinetics of the initial elevation was
steeply dependent on AMD concentration, with maximal increase observed between 10 and 20 ␮M AMD. Elevation of
cytosolic calcium was found to correlate with AMD sensitivity;
thus, the addition of 10 ␮M AMD elicited the largest luminescence changes in the pmr1 mutant, whereas the pmc1 mutant
showed similar changes relative to wild type (Fig. 2D). Estimated cytosolic calcium concentrations of the initial peak,
based on the calibration described under “Experimental Procedures,” reached 0.7 ␮M in wild type and pmc1 and close to 1 ␮M
in pmr1 cells; the second sustained rise remained around 0.5
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
28835
FIG. 3. Contribution of extracellular and vacuolar stores to AMD-induced Ca2ⴙ elevations. A, aequorin luminescence of K601 (WT) and
the isogenic yvc1 mutant was monitored upon the addition of 15 ␮M AMD to cells resuspended at a concentration of 1 OD/ml in SC medium with
or without BAPTA (5 mM), as indicated. B, the addition of 1 mM Al3⫹ or La3⫹ to wild type cells immediately prior to measurement of aequorin
luminescence blocked the initial influx of Ca2⫹ but not the slower rise. All traces were in response to 15 ␮M AMD; the control trace had no ions
added. C, the initial peak of luminescence measured in response to 15 ␮M AMD in pmc1vcx1 and pmc1vcx1yvc1 mutants showed a large (14-fold)
increase in magnitude relative to an isogenic control strain (yvc1).
␮M for WT and pmc1 but stayed at 0.75 ␮M for pmr1.
Ca2⫹ Entry in Response to AMD Comes from Extracellular
and Intracellular Stores—It was of interest to determine the
relative contribution of extracellular and intracellular Ca2⫹
stores to the biphasic elevations of cytosolic Ca2⫹ observed in
Fig. 2. When we used the membrane-impermeant cation chelator BAPTA (5 mM) to deplete extracellular Ca2⫹, the first
peak of AMD-induced aequorin luminescence in wild type cells
was eliminated, indicating that external calcium contributed to
the rapid, initial rise in cytoplasmic calcium (Fig. 3A). Following a short lag, a broad, sustained rise was observed in BAPTAtreated cells that probably corresponded to the second phase of
luminescence seen in the absence of BAPTA. These results
implied that Ca2⫹ release from intracellular stores must contribute to cytosolic elevation of Ca2⫹ and, further, that Ca2⫹
entry from extracellular sources is not required to trigger release from stores. The transient receptor potential-like channel
Yvc1 has been shown to mediate Ca2⫹ release from the vacuole
in response to hypertonic shock (30). We investigated the contribution of vacuolar Ca2⫹ release by examining aequorin luminescence in the yvc1 mutant (Fig. 3A). The changes in intracellular Ca2⫹ in the yvc1 mutant were essentially similar to
wild type in the absence of BAPTA. The first peak of wild type
cells reached ⬃0.8 – 0.9 ␮M, and that of yvc1 was about 0.7 ␮M.
The second sustained rise in both strains ranged from 0.59 to
0.66 ␮M of free calcium. Strikingly, BAPTA treatment completely abolished Ca2⫹ elevation in the yvc1 mutant. These
results indicate that whereas the first peak of AMD-induced
cytosolic Ca2⫹ comes exclusively from outside, the later elevation derives from both extracellular and vacuolar sources.
Courchesne and Ozturk reported that Mid1 could be the
channel through which AMD-induced calcium entry occurs.
They showed that the Ca2⫹ transient in mid1 mutant was
5-fold lower than the wild type control although not abolished
(2). In Fig. 3B, we show that the addition of the calcium channel blockers La3⫹ or Al3⫹ abolished the first elevation of Ca2⫹
but not the second, reaffirming the conclusion from Fig. 3A that
the initial rise in luminescence comes solely from extracellular
Ca2⫹, whereas the slower increase is derived from both extracellular and intracellular stores. High concentrations (10 mM)
of extracellular Ca2⫹ or Mg2⫹ also diminished the initial peak,
possibly due to desensitization/inactivation of the putative
plasma membrane channel (not shown).
Figs. 2 and 3 present the initial elevation of cytosolic Ca2⫹
being rapidly reversed within the first 1–1.5 min after AMD
exposure. In pmr1, the initial rise of calcium is higher, and the
decay is slower (Fig. 2D), suggesting that active Ca2⫹ transport
into the Golgi/secretory pathway by the Pmr1 pump contributes in part to the decay of aequorin luminescence. Since the
vacuole is known to be the major store for cellular calcium, we
28836
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
TABLE I
Genome-wide screen of single gene deletions for AMD-sensitive strains
Almost 5000 mutants, each of which lacks a nonessential gene, were
grown in YPD supplemented with 7 ␮M AMD as described under
“Experimental Procedures.” 36 mutants were identified with enhanced
sensitivity to AMD and classified to distinct functional classes
according to the Saccharomyces Genome Database (SGD, Stanford University).
Mutant types
Mutants
Transporters
Ergosterol biosynthesis
Trafficking pathways
Kinases/phosphatases
Pleiotropic drugsensitive mutants
Transcription factors
Others
Unknowns
pmr1, pdr5, cup5/vma3, vma13, vma1
erg3, erg6, erg24
rcy1, vps65, vps16, vps45, cog6
slt2, ypk1, sac1, opi1, ptc1
lem3, cdc50
swi3, rtf1, rrn10, uga3
pai3, elg1, nbp2
YMR299C, ps110, YHR155W,
YHR035W, rai1, YBL083C, YMR122C,
YDR271C, YOL087C
hypothesized a significant contribution of vacuolar Ca2⫹ sequestration in response to AMD-induced calcium influx. Indeed, we saw that in the absence of both known vacuolar Ca2⫹
transporters, Vcx1 and Pmc1, the first luminescent peak corresponding to extracellular Ca2⫹ uptake was greatly elevated
(Fig. 3C), with an estimated maximum of 3.3 ␮M (concentration
calculation described under “Experimental Procedures”). This
is consistent with a prominent role for vacuolar sequestration
in the Ca2⫹ homeostatic mechanism. The disruption of all three
known Ca2⫹ removal pathways, Pmr1, Pmc1, and Vcx1, is
known to be nonviable (14), most likely due to a complete loss
of cytosolic Ca2⫹ homeostasis.
Large Scale Screen of a Yeast Deletion Library for Amiodarone-sensitive Mutants—The S. cerevisiae gene deletion library
offers a powerful tool for the assignment of new functions to
sequenced genes. Since AMD is a novel antimycotic with unknown targets in yeast, it was of interest to identify additional
genes contributing to AMD sensitivity that might lead to further insight into the cellular mechanism of AMD-mediated
toxicity. We twice screened a library of nearly 5000 single,
nonessential gene deletions for hypersensitivity to AMD (see
“Experimental Procedures”). 15 mutants were identified in the
first screen, and 21 more were found in the second screen. They
were all retested several times to confirm their drug-sensitive
phenotype (Table I). The identification of pmr1 from this genome-wide screen provided independent verification of the results shown in Fig. 1A. We also found pdr5, a mutant lacking
the pleiotropic drug resistance ATPase and widely known to be
hypersensitive to a variety of cytotoxic agents (31). Pdr5 is
likely to be involved in detoxification of AMD by mediating
drug efflux from the cell. Deletion of individual subunits of the
vacuolar H⫹-ATPase (vma1, vma3, and vma13) abolishes the
vacuolar proton electrochemical gradient, which in turn, is
likely to limit vacuolar calcium sequestration. The other isolates fell into clusters that include mutants deficient in components of the ergosterol biosynthesis pathway, intracellular
trafficking pathways, lipid signaling and kinases, transcription
factors, and genes with unknown functions. Some of these
strains have been shown previously to have pleiotropic sensitivity to drugs and may be involved with AMD toxicity in a
more general rather than specific way. For example, lem3 and
cdc50 mutants have deletions in homologous genes of unknown
function and are known to be more sensitive to brefeldin A (32)
and methyl methanesulfonate and hydroxylurea (33), respectively. The ergosterol biosynthesis pathway is the primary
target of many antifungals (34), and disruption of this pathway
damages the integrity of the cell membrane. It was therefore
not surprising to find erg mutants (erg3, erg6, and erg24) with
enhanced sensitivity to AMD. However, the discovery of mutants defective in various components of intracellular trafficking pathways, such as cog6, vps16, vps45, vps65, and rcy1 was
unexpected. We speculate that mutations in trafficking and
signaling pathways might indirectly disrupt Ca2⫹ homeostasis,
resulting in hypersensitivity to AMD.
Based on our results thus far, we have suggested a correlation between AMD sensitivity and increased Ca2⫹ levels in the
cytoplasm. We sought to confirm this hypothesis by examining
calcium-dependent luminescence of aequorin in three representative mutants identified from the genome-wide screen for
AMD hypersensitivity. As shown in Fig. 4, all three mutants
chosen (vma3, vps45, and rcy1) displayed substantially higher
elevations of cytosolic Ca2⫹ upon the addition of AMD despite
the differences in their individual cellular roles. vma3 lacks the
H⫹-transporting subunit c of the vacuolar H⫹-ATPase (35);
vps45 is a vacuolar protein sorting mutant defective in Golgi- or
cytosol-to-vacuole transport (36), and rcy1 is important for
endocytosis/recycling at the plasma membrane (37). In the
vma3 and vps45 mutants, upon the addition of AMD the first
peak of aequorin luminescence is elevated up to 1.2 ␮M and
broadened, suggesting that defects in the function and biogenesis of the vacuole compromise its ability to effectively sequester Ca2⫹ (Fig. 4A). The rcy1 mutant had a unique response to
AMD, showing significant increases in both luminescence
phases such that the first peak, which reached an estimated
Ca2⫹ concentration of 2 ␮M, was not completely abolished before the greatly amplified, second phase occurred (Fig. 4B). To
determine whether the second elevation of calcium in this
mutant came from external sources, we examined aequorin
luminescence in the presence of BAPTA, added to chelate extracellular calcium. As seen in other strains (Fig. 3), we showed
that the first peak of cytosolic Ca2⫹ was abolished upon BAPTA
treatment, but the second, sustained elevation of Ca2⫹ remained (Fig. 4B). Our results suggest AMD-induced release of
Ca2⫹ from internal stores, possibly the vacuole, cannot be effectively cleared from the cytoplasm in the rcy1 mutant. Alternatively, Ca2⫹-filled vesicles derived from the Golgi may require the function of the Rcy1 protein for exocytic unloading.
Synergism between AMD and Azoles—Conventional azole
antifungals, such as miconazole (MIC) and fluconazole (FLUC),
arrest cell growth and are therefore fungistatic rather than
fungicidal. A major problem in the long term treatment with
azoles is the appearance of resistant fungi. AMD has been
reported to have broad range fungicidal effects, particularly
against Cryptococcus (1), although the reported minimal inhibitory concentrations may be too high to be practical for treatment without adverse side effects. We first tested the effect of
low concentrations of AMD (2– 4 ␮M) on cell viability of Saccharomyces in the presence of therapeutic doses of miconazole
(1 ␮M). After 24 h of exposure, neither drug alone had significant effect on cell viability, as determined by methylene blue
staining (Fig. 5A). The synergistic effect of the two drugs was
seen at all concentrations tested and is shown at 4 ␮M AMD;
cell survival fell dramatically from close to 100% in AMD or
MIC alone to less than 10% in the presence of both drugs. To
examine the metabolic state of drug-treated yeast, cells were
stained with the fluorescent dye FUN-1. Without drugs, the
cells were metabolically active and able to process FUN-1 into
cylindrical intravacuolar structures (Fig. 5B). On the other
hand, when grown in the combined presence of AMD (4 ␮M) and
MIC (1 ␮M) for 24 h, nearly all cells turned metabolically
quiescent, and FUN-1 remained a green-yellow, diffused stain
in the cytoplasm (Fig. 5C).
Similar findings were observed using the pathogenic yeasts
C. albicans and C. neoformans (Fig. 6). As reported (1), AMD
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
28837
FIG. 4. AMD-sensitive vma3, vps45, and rcy1 strains cannot effectively clear Ca2ⴙ from the cytosol. Isogenic wild type BY4742 (WT)
and mutant strains were treated with 15 ␮M AMD immediately prior to measurement of aequorin luminescence as described under “Experimental
Procedures.” A, the initial peak of Ca2⫹ is greatly elevated (and its subsequent decay is slowed) in vma3 and vps45 mutants, relative to WT. B, in
the rcy1 mutant, both phases of cytosolic Ca2⫹ entry are greatly elevated, relative to WT. The addition of BAPTA (5 mM) to the cultures immediately
prior to AMD supplement abolished the first peak of Ca2⫹ in both WT and rcy1 but not the second, slower elevation.
FIG. 5. Synergism between amiodarone and miconazole in S.
cerevisiae. A, wild type S. cerevisiae (BY4742) was grown in SC medium in the absence or presence of AMD (4 ␮M), or MIC (1 ␮M), individually or in combination. After 24 h, the cells were stained with
methylene blue and counted as described under “Experimental Procedures.” Viability was calculated as the number of unstained cells divided by the total number of cells counted (about 250 –300), expressed
as a percentage. The average of three separate experiments is shown. B
and C, wild type cells (BY4742) were grown overnight in SC medium
without (B) and with both drugs (C) and then stained with 4 ␮M FUN-1
for 1 h at 30 °C. Final concentrations of AMD and MIC were 4 ␮M and
1 ␮M, respectively. Excitation wavelength was 488 nm, and emission
was at 520 and 600 nm.
inhibited growth of both yeasts, whereas an effective dose of
FLUC (8 or 16 ␮g/ml) completely prevented cell growth (not
shown). Following a 20-h exposure to drugs, cells previously
exposed to low levels of AMD (2 or 4 ␮M) or to FLUC (8 or 16
␮g/ml) alone essentially recovered in drug-free medium, growing to levels approaching control (uninhibited) levels. In contrast, the combination of drugs inhibited regrowth of C. albicans (Fig. 6A) and C. neoformans (Fig. 6B) by 95–98%,
FIG. 6. Synergism between amiodarone and fluconazole in
pathogenic fungi. A, C. albicans SC5314 was grown in 1 ml of YPD
supplemented with 2 ␮M and 4 ␮M AMD in the presence or absence of 8
␮g/ml FLUC at 37 °C. After 20 h, 5 ␮l of cells were diluted into 1.5 ml
of drug-free YPD and grown at 37 °C for another 20 h. Growth (absorbance at 600 nm) was plotted as a percentage of control (no drug). Data
points are averages of duplicates, with S.D. less than 5%. B, C. neoformans JEC21 was grown in 5 ml of YPD at 30 °C, shaken vigorously for
20 h. The medium had 4 ␮M AMD with or without 16 ␮g/ml FLUC. 5 ␮l
of cells exposed to AMD alone or 20 ␮l of cells exposed to both drugs
were diluted into 10 ml of YPD and grown for 24 – 48 h at 30 °C until
saturation. Growth was measured as optical density at 600 nm and
then plotted as a percentage of control (no drug). Data points are
averages of duplicates, with S.D. equal to or less than 10%.
indicative of an effective loss of viability. It should be noted
that therapeutic plasma levels of 4 ␮M have been reported for
AMD and its metabolite, desethylamiodarone (1), suggesting
that these clinically relevant doses may be additionally effective in antifungal therapy.
DISCUSSION
AMD is an important therapeutic agent used in the treatment of life-threatening ventricular arrhythmias and supraventricular dysrhythmias (38). AMD and its active metabolite, desethylamiodarone, are unusual in their long biological
half-life (25–110 days) and their broad range of targets. In
cardiac cells, AMD inhibits ion channels such as L-type Ca2⫹
channels and Na⫹ and K⫹ channels, thus prolonging the duration of action potential (39). Although the therapeutic benefits of AMD are well established, toxic side effects occur in
28838
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
noncardiac tissues, most prominently the lungs and the thyroid
gland. There is evidence that AMD causes apoptosis in thyrocytes (40) and pneumocytes (41). Alveolar macrophages treated
with AMD show defects in endocytosis at the steps after the
late endosome and missort lysosomal enzymes to the plasma
membrane (42). Of particular interest to this study are the
reports of acute AMD exposure leading to a sustained cytoplasmic Ca2⫹ rise in synaptosomes (43) and in aortic or pulmonary
endothelial cells (44, 45). Our work demonstrates a correlation
between AMD toxicity and increased cytoplasmic Ca2⫹ levels in
S. cerevisiae, exemplifying the conservation of Ca2⫹ transport
and signaling pathways between yeast and higher eukaryotes.
Therefore, a better understanding of the cellular basis of AMD
toxicity in this simple model organism is relevant to understanding and eventually circumventing the toxic side effects of
the drug in humans.
We show that low, micromolar doses of AMD, in the range
where growth inhibition is observed, stimulate Ca2⫹ uptake
through a plasma membrane channel that is inhibited by Ca2⫹
channel blockers such as Al3⫹ and La3⫹ and by competition
with Mg2⫹ ions. Courchesne and Ozturk (2) have suggested
that one pathway for Ca2⫹ entry is the stretch-activated channel Mid1. They found that the mid1 mutant was less sensitive
to AMD than cch1, suggesting that loss of Cch1 allows Mid1 to
function independently. However, we find that the mid1 mutant exhibits enhanced growth sensitivity to AMD similar to
that of the cch1 mutant (not shown). The reason for this discrepancy is not clear and warrants further studies.
We provide new evidence for Ca2⫹ release from the vacuole
via the transient receptor potential-like channel, Yvc1. Taken
together with a previously published report of the role of this
channel in mediating the response to hypertonic shock (30), it
would appear that Yvc1 is the major release mechanism for
intracellular stores of Ca2⫹ in yeast. Whereas depletion of
extracellular Ca2⫹ or the addition of Ca2⫹ channel blockers
abolished the immediate elevation of cytosolic Ca2⫹, they did
not prevent the Yvc1-mediated release from vacuolar stores,
which continued to increase during the time course of the
assay. Based on our calculations of calcium concentration, the
two phases in WT cells were estimated to correspond to 0.7 and
0.5 ␮M free Ca2⫹, respectively, in response to 10 ␮M AMD. In
strains hypersensitive to AMD, such as rcy1, these values
increased to 2 ␮M and 1.6 –2.5 ␮M, respectively. We speculate
that it is the second sustained rise of cytoplasmic Ca2⫹ that
eventually becomes toxic to AMD-treated cells. Given that the
sequence of calcium entry from extracellular sources clearly
precedes that from intracellular stores in our experiments, we
can rule out a role for capacitative calcium entry in the immediate response to AMD, although such mechanisms are likely to
contribute to the sustained phase of calcium increase.
It is well known that excessive cytosolic Ca2⫹ is toxic (46),
and indeed, we show a consistent correlation between hypersensitivity to AMD and excessive elevation of cytosolic Ca2⫹.
For example, in the absence of the Golgi Ca2⫹ pump, Pmr1,
elevated Ca2⫹ levels in response to AMD result in a hypersensitive phenotype, consistent with the important role of this
pump in the dynamic regulation of cellular Ca2⫹ levels. Among
the deletion strains that we analyzed, only the pmc1 mutant
was moderately tolerant to AMD and was seen to have similar
elevations of cytosolic Ca2⫹, relative to wild type. Paradoxically, very high levels (10 mM) of extracellular Ca2⫹ or Mg2⫹
ameliorated the adverse effects of AMD (1) (Fig. 2A), seemingly
in contradiction to our hypothesis. A possible explanation is
that the Ca2⫹ influx pathway(s), probably a Ca2⫹ channel, is
down-regulated in the presence of excess amounts of extracellular Ca2⫹. A similar response has been observed in Nicotiana
cells exposed to Phytophthora toxins (10), in which calcium
influx first increased and then decreased with increasing extracellular Ca2⫹ levels.
In order to gain insight into the molecular target(s) of AMD
in yeast, we have defined distinct cellular pathways and components that confer drug sensitivity upon disruption (Table I).
The genome-wide screens were not exhaustive, and it is possible that many other mutants were missed. However, our results are promising, since we have identified several interesting mutants (e.g. erg3, rcy1, vps45) that, with further studies,
would shed more light on the molecular basis of AMD toxicity
as well as calcium homeostasis in fungi. The enhanced sensitivity of trafficking mutants may be due to defects in drug
detoxification and/or Ca2⫹ clearance via the secretory pathway
or vacuolar sequestration. For example, the rcy1 mutant is
defective in exit from an early, sorting endosome which has
been postulated to receive traffic from the secretory and endocytic pathways (37). AMD toxicity in the rcy1 mutant might be
due to defective traffic of Golgi- or vacuole-derived vesicles
loaded with Ca2⫹ or in the recycling of endocytosed AMD to the
plasma membrane. On the other hand, vps45, a Golgi- or cytosol-to-vacuole transport mutant (36), and vma3, a strain lacking the proton-transporting subunit c of the vacuolar H⫹ATPase (35), are both defective in vacuolar function and
biogenesis, so that Ca2⫹ cannot be effectively sequestered. The
drug sensitivity observed in signaling mutants like sac1 and
ptc1 is also interesting and unexpected. SAC1 encodes a phosphoinositide phosphatase, in the absence of which the cell
contains an irregularly shaped vacuole surrounded by lipid
droplets (47), suggesting that vacuolar function is compromised
in this mutant. Ptc1 is a negative regulator of the high osmolarity glycerol pathway and has been shown to maintain the
basal kinase activity of Hog1 (48). Clearly, our findings warrant further investigation to elucidate the relationships between AMD toxicity and PTC1 as well as other genes identified
in the screen.
It remains to be determined if AMD directly binds and affects a Ca2⫹ channel or disrupts an intracellular signaling
pathway that then leads to cytosolic Ca2⫹ entry. We note that
in addition to its action on ion channels, other reported cellular
activities of AMD and desethylamiodarone encompass binding
and activation of heterotrimeric G proteins (49), inhibition of
phospholipases A1 and A2 (50), and binding and inhibition of
calmodulin (51, 52) as well as protein kinase C (53). In any
case, our work suggests the possibility of using AMD as a novel,
broad range antimycotic that is likely to be effective against
even drug-resistant pathogenic fungi. Most, if not all, of the
cellular pathways and components identified in our screen for
mutants hypersensitive to this drug (Table I) are expected to be
conserved in the pathogenic counterparts of bakers’ yeast. Mucosal and systemic fungal infections are among the main killers
in immunocompromised patients, and resistant fungi have become increasingly prevalent. Conventional drugs such as the
azoles are only fungistatic, whereas others that are fungicidal,
such as the polyenes, have harmful side effects that preclude
long term treatment (54). The target of both azoles and polyenes is ergosterol and its biosynthesis pathway. One mechanism of drug resistance is the inactivation of Erg3, a sterol
desaturase, which results in the accumulation of 14␣-methylfecosterol, a growth-promoting sterol (55). In the absence of
ergosterol, the erg3 mutant is resistant to both amphotericin B
(a polyene) and azoles. Interestingly, the erg3 mutant was
identified in our screen for AMD sensitivity (Table I), suggesting that cytotoxicity of AMD occurs by a mechanism distinct
from that of the conventional antifungals. If added to an antifungal regimen, AMD could potentially enhance the efficacy by
Amiodarone Disrupts Ca2⫹ Homeostasis in Yeast
eliminating cells whose growth is only halted by the other
fungistatic drugs. In combination, specificity against pathogenic fungi would be conferred by the azoles, and toxicity would
be conferred by AMD. To be useful therapeutically, the low
concentrations of AMD that are synergistic with azoles should
have minimal toxicity to humans.
Although the strategy of combining different classes of antifungals is theoretically attractive, some drugs, when used concomitantly or sequentially, exert no synergism and may even
be antagonistic. Specifically, A. fumigatus, when pretreated
with noninhibitory doses of itraconazole (an azole), developed
resistance to amphotericin B (a polyene) (56). A combination of
fluconazole (an azole) and caspofungin or anidulafungin (two
echinocandins) produced no significant changes in efficacy of
individual drugs, although the former targets the ergosterol
biosynthesis pathway, whereas the latter two inhibit cell wall
synthesis (54). Therefore, the strong synergism between miconazole or fluconazole and AMD demonstrated in this work promises a potentially effective regimen against severe and lifethreatening fungal infections.
Acknowledgments—S. S.G. thanks Dr. C. M. C. Allen and Dr. R.
Jones of Lucy Cavendish College for making the elective research
possible. We thank Dr. B. Cormack (The Johns Hopkins University
School of Medicine) for the kind gifts of fluconazole and C. albicans and
Dr. J. Heitman (Duke University Medical Center) for C. neoformans.
REFERENCES
1. Courchesne, W. E. (2002) J. Pharmacol. Exp. Ther. 300, 195–199
2. Courchesne, W. E., and Ozturk, S. (2003) Mol. Microbiol. 47, 223–234
3. Cruz, M. C., Goldstein, A. L., Blankenship, J. R., Del Poeta, M., Davis, D.,
Cardenas, M. E., Perfect, J. R., McCusker, J. H., and Heitman, J. (2002)
EMBO J. 21, 546 –559
4. Matsumoto, T. K., Ellsmore, A. J., Cessna, S. G., Low, P. S., Pardo, J. M.,
Bressan, R. A., and Hasegawa, P. M. (2002) J. Biol. Chem. 277,
33075–33080
5. Bonilla, M. Nastase, K. K., and Cunningham, K. W. (2002) EMBO J. 21,
2343–2353
6. Edlind, T., Smith, L., Henry, K., Katiyar, S., and Nickels, J. (2002) Mol.
Microbiol. 46, 257–268
7. Locke, E. G., Bonilla, M., Liang, L., Takita, Y., and Cunningham, K. W. (2000)
Mol. Cell. Biol. 20, 6686 – 6694
8. Andjelic, S., Khanna, A., Suthanthiran, M., and Nikolic-Zugic, J. (1997) J. Immunol. 158, 2527–2534
9. Kurzweilova, H., and Sigler, K. (1993) Folia Microbiol. 38, 524 –526
10. Lecourieux, D., Mazars, C., Pauly, N., Ranjeva, R., and Pugin, A. (2002) Plant
Cell 14, 2627–2641
11. Sorin, A., Rosas, G., and Rao, R. (1997) J. Biol. Cell 272, 9895–9901
12. Marchi, V., Sorin, A. Wei, Y., and Rao, R. (1999) FEBS Lett. 454, 181–186
13. Ton, V.-K., Mandal, D., Vahadji, C., and Rao, R. (2002) J. Biol. Chem. 277,
6422– 6427
14. Cunningham, K. W., and Fink, G. R. (1996) Mol. Cell. Biol. 16, 2226 –2237
15. Cunningham, K. W., and Fink, G. R. (1994) J. Cell Biol. 124, 351–363
16. Pozos, T. C., Sekler, I., and Cyert, M. S. (1996) Mol. Cell. Biol. 16, 3730 –3741
17. Miseta, A., Kellermayer, R., Aiello, D. P., Fu, L., and Bedwell, D. M. (1999)
FEBS Lett. 451, 132–136
18. Palmer, C. P., Zhou, X. L., Lin, J., Loukin, S. H., Kung, C., and Saimi, Y. (2001)
Proc. Natl. Acad. Sci. U. S. A. 98, 7801–7805
28839
19. Paidhungat, M., and Garrett, S. (1997) Mol. Cell. Biol. 17, 6339 – 6347
20. Muller, E. M., Locke, E. G., and Cunningham, K. W. (2001) Genetics 159,
1527–1538
21. Cyert, M. S. (2001) Annu. Rev. Genet. 35, 647– 672
22. Fox, D. S., and Heitman, J. (2002) Bioessays 24, 894 –903
23. Mandal, D., Woolf, T., and Rao, R. (2000) J. Biol. Cell 273, 23933–23938
24. Wei, Y., Chen, J., Rosas, G., Tompkins, D. A., Holt, A., and Rao, R. (2000)
J. Biol. Chem. 275, 23927–23932
25. Fonzi, W. A., and Irwin, M. Y. (1993) Genetics 134, 717–728
26. Cruz, M. C., Goldstein, A. L., Blankenship, J., Poeta, M. D., Perfect, J. R.,
McCusker, J. M., Bennani, Y. L., Cardenas, M. E., and Heitman, J. (2001)
Antimicrob. Agents. Chemother. 45, 3162–3170
27. Allen, D. G., Blinks, J. R., and Prendergast, F. G. (1977) Science 195, 996 –998
28. Gage, M. J., Bruenn, J., Fischer, M., Sanders, D., and Smith, T. J. (2001) Mol.
Microbiol. 41, 775–785
29. Durr, G., Strayle, J., Plemper, R., Elbs, S., Klee, S. K., Catty, P., Wolf, D. H.,
and Rudolph, H. K. (1998) Mol. Biol. Cell 9, 1149 –1162
30. Denis, V., and Cyert, M. S. (2002) J. Cell Biol. 156, 29 –34
31. Bauer, B. E., Wolfger, H., and Kuchler, K. (1999) Biochim. Biophys. Acta. 1461,
217–236
32. Sitcheran, R., Emter, R., Kralli, A., and Yamamoto, K. R. (2000) Genetics 156,
963–972
33. Moir, D., Stewart, S. E., Osmond, B. C., and Botstein, D. (2000) Genetics 100,
547–563
34. Tkacz, S., and DiDomenico, B. (2001) Curr. Opin. Microbiol. 4, 540 –545
35. Eide, D. J., Bridgham, J. T., Zhao, Z., and Mattoon, J. R. (1993) Mol. Gen.
Genet. 241, 447– 456
36. Abeliovich, H., Darsow, T., and Emr, S. D. (1999) EMBO J. 18, 6005– 6016
37. Wiederkehr, A., Avaro, S., Prescianotto-Baschong, C., Haguenauer-Tsapis, R.,
and Riezman, H. (2000) J. Cell Biol. 149, 397– 410
38. Stevenson, W. G., Ellison, K. E., Sweeney, M. O., Epstein, L. M., and Maisel,
W. H. (2002) Cardiol. Rev. 10, 8 –14
39. Kodama, I., Kamiya, K., and Toyama, J. (1999) Am. J. Cardiol. 84, 20R–28R
40. Di Matola, T., D’Ascoli, F., Fenzi, G., Rossi, G., Martino, E., Bogazzi, F., and
Vitale, M. (2000) J. Clin. Endocrinol. Metab. 85, 4323– 4330
41. Bargout, R., Jankov, A., Dincer, E., Wang, R., Komodromos, T., Ibarra-Sunga,
O., Filippatos, G., and Uhal, B. D. (2000) Am. J. Physiol. 278, L1039 –L1044
42. Baritussio, A., Marzini, S., Agostini, M., Alberti, A., Cimenti, C., Bruttomesso,
D., Manzato, E., Quaglino, D., and Pettenazzo, A. (2001) Am. J. Physiol.
281, L1189 –L1199
43. Kodavanti, P. R., Pentyala, S. N., Yallapragada, P. R., and Desaiah, D. (1992)
Naunyn Schmiedeberg’s Arch. Pharmacol. 345, 213–221
44. Himmel, H. M., Dobrev, D., Grossmann, M., and Ravens, U. (2000) Naunyn
Schmiedeberg’s Arch. Pharmacol. 362, 489 – 496
45. Powis, G., Olsen, R., Standing, J. E., Kachel, D., and Martin, W. J., II (1990)
Toxicol. Appl. Pharmacol. 103, 156 –164
46. McConkey, D. J., and Orrenius, S. (1997) Biochem. Biophys. Res. Commun.
239, 357–366
47. Foti, M., Audhya, A., and Emr, S. D. (2001) Mol. Cell. Biol. 12, 2396 –2411
48. Warmka, J., Hanneman, J., Lee, J., Amin, D., and Ota, I. (2001) Mol. Cell. Biol.
21, 51– 60
49. Hageluken, A., Nurnberg, B., Harhammer, R., Grunbaum, L., Schunack, W.,
and Seifert, R. (1995) Mol. Pharmacol. 47, 234 –240
50. Honegger, U. E., Zuehlke, R. D., Scuntaro, I., Schaefer, M. H., Toplak, H., and
Wiesmann, U. N. (1993) Biochem. Pharmacol. 45, 349 –356
51. Deziel, M. R., Davis, P. J., Davis, F. B., Cody, V., Galindo, J., Jr., and Blas,
S. D. (1989) Arch. Biochem. Biophys. 274, 463– 470
52. Vig, P. J., Yallapragada, P. R., Kodavanti, P. R., and Desaiah, D. (1991)
Pharmacol. Toxicol. 68, 26 –33
53. Vig, P. J., and Desaiah, D. (1991) Neurotoxicology 12, 595– 601
54. Roling, E. E., Klepser, M. E., Wasson, A., Lewis, R. E., Ernst, E. J., and Pfaller,
M. A. (2002) Diagn. Microbiol. Infect. Dis. 43, 13–17
55. Lupetti, A., Danesi, R., Campa, M., Del Tacca, M., and Kelly, S. (2002) Trends
Mol. Med. 8, 76 – 81
56. Kontoyiannis, D. P., Lewis, R. E., Sagar, N., May, G., Prince, R. A., and
Rolston, K. V. (2000) Antimicrob. Agents. Chemother. 44, 2915–2918